I've been working on a graphics programming tool a...
# of-graphics
a
I've been working on a graphics programming tool aims to do just that; deconstructing every part of a graphics pipeline, and giving artists all that power in a visual interface. Main difference between most visual programming tools is that we're relying on standard APIs (like say OpenVDB or any other python/c++ library) and simply creating a visual layer (that has an extensive tagging system to make it easy as hell to find anything you might want) that anyone can interact with. Your result can be previewed at runtime, converted to code or both. Still quite early in the dev phase, but hopefully will have more shiny stuff to share on this project later this year 🙂
❤️ 4
i
Do you have a demo video of this tool in action?
Or is it too early for that?
a
Yep a tad early 😁
Long term goal is to be able to generate tools like this in no time:

https://i.imgur.com/VCWHqUD.gif

(this is a Maya plugin that I did some 5 years ago)
💯 1
i
Cool!
s
@Adnan Chaumette yeah I think one of the things that's missing in most visual tools for artists is the ability to easily create new visual workflows\tools\etc in a visual editor. Working on game teams I found that engineering (aka me) was often the bottleneck for relatively minor UX and workflow changes
I don't think its a trivial problem
and, you're right, in a lot of environments it can be solved with text scripting
Unreal 4 kind of allows for some of it also, but I haven't worked on a AAA UE4 project with a big team so I'm not sure what that looks like
a
Since they've added python support, UE4 has become very flexible actually, and all 3D softwares also have python support for their C++ APIs as well. In 2016 I basically started publishing really small scripts that improved specific parts of Autodesk Maya, and before I knew it I had over 30K paying users and that's where I really understood that you don't even have to create new frameworks, just give people a better way to interact with existing APIs. Of course there are always limits to said APIs, which is why we're building C++ libraries (with python bindings) in parallel for more advanced geometry and image processing.
👍 1
In the short term the goal is to give people a new way of interacting with code, in the long-term our ecosystem will leverage more APIs built specifically around certain issues that we want to solve. And the fact that the end result is always standard code means that there's no lock-in to our platform; you can make something you like, share it with your friends and voilà. We're still a fairly small team (that's hiring CG programmers, btw!) but have secured the funding/structure to make the most out of this project, and hopefully we'll get to share more about it soon! 🙂